LUND UNIVERSITY SCHOOL OF ECONOMICS AND MANAGEMENT MASTER OF FINANCE STOCK LIQUIDITY AS A DETERMINANT OF CREDIT DEFAULT SWAP SPREADS MASTER THESIS IN FINANCE Authors: Mehmet Caglar Kaya, Radu-Dragomir Manac Supervisors: Prof. Hossein Asgharian, Prof. Ola Bengtsson May 2013
79
Embed
STOCK LIQUIDITY AS A DETERMINANT OF CREDIT DEFAULT SWAP ...lup.lub.lu.se/student-papers/record/3808236/file/3808240.pdf · stock liquidity as a determinant of credit default swap
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
LUND UNIVERSITY
SCHOOL OF ECONOMICS AND MANAGEMENT
MASTER OF FINANCE
STOCK LIQUIDITY AS A DETERMINANT OF CREDIT DEFAULT SWAP SPREADS
MASTER THESIS IN FINANCE
Authors:
Mehmet Caglar Kaya, Radu-Dragomir Manac
Supervisors:
Prof. Hossein Asgharian, Prof. Ola Bengtsson
May 2013
1
ABSTRACT
This research investigates the effect of stock liquidity on credit default swap spreads. The
relationship between stock liquidity and CDS spreads is tested empirically using a panel data
of 82 companies spanning a period of 64 months. To ensure the accuracy of our findings, we
use three proxies for stock liquidity, namely the bid-ask spread, Amihud illiquidity measure
and the turnover ratio. When controlling for other known firm-level factors, we obtain a
relationship between stock liquidity measures and CDS spreads, indicating that higher stock
liquidity leads to lower CDS spread. This relation also holds when macroeconomic factors are
used as control variables. Thereby, we manage to find a link between the stock market and the
CDS market. This relationship helps predict the movement of CDS spreads by analyzing
3.1 Data .............................................................................................................................................. 21
3.1.1. CDS Data .............................................................................................................................. 21
3.1.2. Stock Liquidity Data ............................................................................................................. 22
3.1.3. Control Variables Data ......................................................................................................... 24
3.2. Model .......................................................................................................................................... 29
The table reports the mean, the median, the maximum value (max.), the minimum value (min.), the standard deviation (std. dev.), the skewness, the kurtosis and the number of observations (N) of the dependent variable (lncds) and the explanatory variables over the period December 2007 to March 2013. The CDS Spread is defined as the natural logarithm of the monthly average mid CDS spread quote for CDS with 5-year maturity. The Bid-Ask Spread is the monthly average of the ratio between the absolute difference of the ask quote and the bid quote divided by their average. The Amihud Illiquidity measure is the monthly average of the daily ratio between the absolute stock return and the euro trading volume. The Turnover Ratio is measured as the monthly average of the ratio between the volume of shares traded and the total number of shares outstanding. The Return is calculated as the monthly average of the natural logarithm of the ratio between the stock price in day t+1 and the stock price in day t. The Volatility is measured as the standard deviation of the stock returns over a month. The Equity Value is calculated as stock price multiplied by the number of shares and is collected directly from Datastream. The Operating Profit Margin is calculated as operating income divided by sales revenue. The Debt to Equity Leverage Ratio is calculated as the ratio of the sum of long term debt, short term debt and current portion of long term debt divided by common equity, in percentage terms. The Debt to Total Assets Leverage Ratio is the ratio of the sum of long term debt, short term debt and current portion of long term debt divided by the sum of total equity, short term debt and current portion of long term debt, in percentage terms. The Interest Coverage Ratio is measured as earnings before interest and tax (EBIT) divided by interest expense, in percentage terms. The GDP Growth Rate is calculated on a monthly basis using the simple return of the nominal GDP of the countries in which the companies in the sample are headquartered. The one-year Interbank Interest Rate is collected with a monthly frequency from Datastream. The five-year and ten-year Government Bond Yields are collected on a monthly frequency from the European Central Bank and national central banks of the countries in which the companies in the sample are headquartered.
29
3.2. Model
Our data is a longitudinal (panel) data in which we analyze firms from different
industries in the cross section dimension throughout time. Based on the literature, using panel
data has some advantages. First, as Kennedy (2009) argues, panel data deals with
heterogeneity in cross-sectional units in each time period thus coping with the omitted
variables problem. Omitting time series variables, which influences the behavior of firm-
specific variables, causes bias in the estimation. Second, panel data provides more variability
through combining variation across cross-sectional units with variation over time, eliminating
any multicollinearity problems to some extent (Kennedy, 2009, 281-282). Third, panel data is
more informative than cross sectional data. It enables us to examine and address how an
independent variable behaves across firms over different months and explain its effects on the
dependent variable. Fourth, panel data lets us perform a better analysis of dynamic
adjustments: cross-sectional data standalone is not able to provide us dynamics, whereas we
would need very lengthy times-series data to obtain better estimates of dynamic behavior
(Kennedy, 2009, 281-282). We analyze 82 firms in the cross-sectional dimension over 64
months and have data for all variables in all firms and all periods. Therefore, we have a
balanced panel data consisting of 5248 observations for each variable.
Before proceeding with the estimation of the model, we first search for
multicollinearity between the independent variables. Multicollinearity refers to the situation
where the explanatory variables are correlated (Gujarati & Porter, 2010). As a consequence of
multicollinearity, “the ordinary least squares (OLS) estimating procedure is not given enough
independent variation in a variable to calculate with confidence the effect it has on the
independent variable” (Kennedy, 2009:193). The correlation matrix presented in Table 2
reports two correlations higher than 0.8. These are the correlations between the five-year bond
yield (fiveyld) and ten-year bond yield (tenyld) which is equal to 0.96 and between the five-
year bond yield (fiveylvd) and one-year interbank rate (intbankr) which is equal to 0.81.
However, these three measures are all proxies for the same variable, namely risk-free rate.
Thus, they are not introduced simultaneously in the same regression throughout the analysis.
Regarding the other variables, the correlation table shows that none are significantly
correlated with each other, pairwise correlations not exceeding +/- 54%.
30
TABLE 2: Correlation Matrix
The table shows the correlations between monthly observations of the CDS spreads (lncds), the stock liquidity proxies, the firm-level control variables and the macroeconomic control variables. The stock liquidity proxies that we use are: the bid-ask spread (bas), Amihud illiquidity measure (illiq) and turnover ratio (trnvr). The firm-level variables are: stock return (r), equity value (lne), volatility (vol), operating margin (opm), debt to equity leverage (levde), debt to total assets leverage (levdta), interest coverage ratio (levintcov). The macroeconomic variables are: GDP growth rate (gdp), one-year interbank interest rate (intbankr), five-year bond yield (fiveyld), ten-year bond yield (tenyld). Data covers the period between December 2007 and March 2013. We note that the only correlations exceeding +/- 0.6 are between the three risk-free rate proxies: the one-year interbank rate and the five and ten year bond yields. Correlations among other variables do not exceed +-0.54, indicating that we do not have multicollinearity in the data set.
In order to look for initial signs of heterogeneity, to have a benchmark result for
comparison purposes and to get an indication if any other potential problems occur, the
pooled regression is run first. The equation for the pooled regression is shown below:
𝑙𝑛𝑐𝑑𝑠𝑖,𝑚= 𝛼𝑖 + 𝛽𝑖𝑙𝑖𝑞 x 𝑙𝑖𝑞𝑖,𝑚+𝛽𝑖𝑟 x 𝑟𝑖,𝑚 + 𝛽𝑖𝑒 x 𝑙𝑛𝑒𝑖,𝑚 +𝛽𝑖𝑣𝑜𝑙 x 𝑣𝑜𝑙𝑖,𝑚 +𝛽𝑖
𝑜𝑝𝑚 x 𝑜𝑝𝑚𝑖,𝑚 +𝛽𝑖𝑙𝑒𝑣 x 𝑙𝑒𝑣𝑖,𝑚 + 𝜀𝑖,𝑚
where we denote the CDS spread of firm i in month m in natural logarithm as 𝑙𝑛𝑐𝑑𝑠𝑖,𝑚 ; the
intercept of the equation as 𝛼𝑖 ; the stock liquidity of firm i in month m as 𝑙𝑖𝑞𝑖,𝑚 ; the
monthly return of stock of firm i in month m as 𝑟𝑖,𝑚 ; the natural logarithm of market value
of equity of firm i at the end of month m as 𝑙𝑛𝑒𝑖,𝑚; the volatility of stock return of firm i in
0,1287 0,6475 0,6475 F-Statistic 258,0934 112,9133 112,9133 Prob. (F-statistic) 0,0000 0,0000 0,0000 *The numbers in brackets represent standard errors for each coefficient.
Table 13 reports the coefficient, standard errors and probabilities for the intercept (c), the bid-ask spread stock
liquidity measure (bas) and the macro-level control variables: GDP growth (gdp) and five-year bond rate
(fiveyld) in three different regressions: the pooled regression using ordinary standard errors, the regression with
cross-section fixed effects using ordinary standard errors and the regression with cross-section fixed effects
using White cross-section standard errors. The R-squared, F-statistic, probability of F-statistic are also reported.
The regressions in Table 13 are different to those in Table 12 only in respect of the proxy for risk-free rate. In
this case, we use the five-year bond yield. We note that all variables, have expected significant relationships with
the dependent variable, CDS spreads.
50
TABLE 14: Regressions of Amihud Stock Liquidity with Macro-level Control Variables; Pooled and Fixed Effects using Ordinary SE and White SE Regressions
REGRESSION
Pooled
Fixed Cross-Section Effect & No Period
Fixed Effect
Fixed Cross-Section Effect & No Period
Fixed Effect
Explanatory Variables Ordinary SE Ordinary SE White Cross Section SE
Ericsson, J., Jacobs, K Oviedo, R. “The Determinants of Credit Default Swap Premia”,
2005, Journal of Financial and Quantitative Analysis, forthcoming;
Ericsson, J., Renault, O. “Liquidity and Credit Risk”, 2002, EFA 2003 Glasgow
Working Paper Series;
Fung, H-G., Sierra, G. E., Yau, J., Zhang G. “Are the US Stock Market and Credit
Default Swap Market Related? Evidence from CDX Indices”, summer 2008, Journal of
Alternative Investments;
Goldstein, M., Hotchkiss, E., Sirri, E. “Transparency and Liquidity: A Controlled
Experiment on Corporate Bonds”, 2006, Working Paper, Boston College;
Gujarati, D. N., Porter D. C. “Essentials of Econometrics”, 2010, 4th Edition, McGraw
Hill Publication, Singapore;
Gunther, J.W., Levonian, M. E., Moore, R. R. “Can the Stock Market Tell Bank
Supervisors Anything They Don’t Already Know?”, 2001, Federal Reserve Bank of
Dallas Economic and Financial Review;
Hsiao, C. “Analysis of Panel Data”, 2005, Cambridge University Press, 2nd edition,
Cambridge;
54
Hull, J. “Options, Futures and Other Derivatives”, 2003, 5th edition, Prentice Hall;
Hull, J., Nelken, I., White, A. “Merton’s Model, Credit Risk and Volatility Skews”,
2004, Journal of Credit Risk, Vol.1, No. 1, pp. 3-28;
Hull, J., Predescu, M., White, A. “The Relationship between Credit Default Swap
Spreads, Bond Yields and Credit Rating Announcements”, 2004, Journal of Banking
and Finance, Vol. 28, Issue 11, pp. 2789-2811;
Irvine, P.J., Pontiff J. “Idiosyncratic Return Volatility, Cash Flows and Product Market
Competition”, 2009, The Review of Financial Studies, Vol. 22, No.3, pp.1149-1177;
Jarrow, R. “Problems with using CDS to imply default probabilities”, 2012, Journal of
Fixed Income Vol.21, Issue 4, pp. 6-12;
Kapar, B., Olmo, J. “The Determinants of Credit Default Swap Spreads in the Presence
of Structural Breaks and Counterparty Risk”, 2011, City University London,
Department of Economics Discussion Paper Series No. 11/02;
Kennedy, P. “A Guide to Econometrics”, 2009, 6th Edition, Blackwell Publishing,
Oxford UK;
Knaup, M., Wagner, W. “A Market-Based Measure of Credit Quality and Banks’
Performance during the Subprime Crisis”, 2008, European Banking Centre Discussion
Paper No 2009-06S, Tilburg University;
Kwan, S. “Firm-specific Information and the Correlation between Individual Stock and
Bonds”, 1996, Journal of Financial Economics, Vol. 40, pp. 63-80;
Meng, L., Gwilym, O. A. “The Determinants of CDS bid-ask spreads”, 2008, Journal of
Derivatives, Vol. 16, No. 1, pp. 70-80;
Mengle, D. “Credit Derivatives: An Overview”, 2007, Federal Reserve Bank of Atlanta,
presented at Atlanta Fed’s Financial Markets Conference “Credit Derivatives: Where’s
the Risk?”;
Merton, R.C. “On the Pricing of Corporate Debt: The Risk Structure of Interest Rates”,
1974, Journal of Finance, Vol. 29, pp. 449-470;
Norden, L. “Credit Derivatives, Corporate News and Credit Ratings”, 2008, University
of Mannheim, Deutsche Bundesbank;
55
Norden, L., Weber, M. “The Comovement of Credit Default Swap, Bond and Stock
Markets: An Empirical Analysis”, 2004, Centre for Financial Studies, Working Paper
No 2004/20, Goethe University, Frankfurt;
Opler, T., Titman, S. “Financial Distress and Corporate Performance”, 1994, Journal of
Finance, Vol. 549, pp. 1015-1040;
Rauh, J.D., Sufi, A. “Capital Structure and Debt Structure”, 2008, National Bureau of
Economic Research, Working Paper 14488;
Rösch, C. G., Kaserer, C. “Market Liquidity in the Financial Crisis: The Role of
Liquidity Commonality and Flight-to-quality”, 2012, Technische Universität Munchen,
Working Paper Series;
Spiegel, M., Wang, X. “Cross-sectional Variation in Stock Returns: Liquidity and
Idiosyncratic Risk”, Yale School of Management, International Center for Finance,
Yale University, Working Paper No. 05-13, EFA 2005 Moscow Meetings Paper;
Suh, C.B., Lee, Y. “Determinants of Banks’ CDS Spreads and Policy Implications”,
2011, Seoul Journal of Economics, Vol. 24, No.4, pp. 575-591;
Tang, D.Y., Yan, H. “Liquidity and Credit Default Swap Spreads”, 2008 EFA Athens
Meetings Paper, 2007 AFA Chicago Meetings Paper;
Tang, D.Y., Yan, H. “Market Conditions, Default Risk and Credit Spreads”, 2010,
Journal of Banking and Finance, Vol. 34, pp. 743-753;
Tang, D.Y., Yan, H. “What Moves CDS Spreads”, 2012, University of Hong Kong
Research Paper;
The Board of the International Organization of Securities Commission “The Credit
Default Swap Market Report”, 2012, FR05/12;
Vassalou, M., Xing, Y. “Default Risk in Equity Returns”, 2004, The Journal of Finance,
Vol. 59, No. 2, pp. 831-868;
Zhu, H. “An Empirical Comparison of Credit Spreads Between the Bond Market and
the Credit Default Swap Market”, 2006, Journal of Financial Services Research, Vol.
29, No. 3, pp. 211-235.
56
7. APPENDIX: 1-) POOLED REGRESSION OF BID-ASK SPREAD STOCK LIQUIDITY WITH FIRM-LEVEL CONTROL VARIABLES
The table depicts the results of the pooled regression with the bid-ask spread (bas) as a proxy for stock liquidity and debt to equity ratio (levde) as a proxy for the leverage control variable. The return (r), equity value (lne), volatility (vol), operating margin (opm) and debt to equity leverage (levde) are the firm-level control variables used in this regression. We use ordinary standard errors. The time frame of the data is December 2007 until March 2013. Bid-ask spread has an insignificant relationship with CDS spread in the pooled regression.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities
2-) RESIDUAL GRAPH OF POOLED REGRESSION WITH FIRM-LEVEL CONTROL VARIABLES
The graph shows the distribution of the residuals of the pooled regression with firm-level control variables. The proxy used for stock liquidity is the bid-ask spread (bas). The presence of heteroskedasticity is observed from the graph.
-1.6
-1.2
-0.8
-0.4
0.0
0.4
0.8
1.2
1.6
2.0
500 1000 1500 2000 2500 3000 3500 4000 4500 5000
LNCDS Residuals
57
3-) REGRESSION OF BID-ASK SPREAD STOCK LIQUIDITY WITH FIRM-LEVEL CONTROLS USING CROSS-SECTION FIXED EFFECTS
The table shows the results of the regression using cross-section fixed effects, which includes the bid-ask spread (bas) as a proxy for stock liquidity and debt to equity ratio (levde) as a proxy for the leverage control variable. The return (r), equity value (lne), volatility (vol), operating margin (opm) and debt to equity ratio (levde) are the firm-level control variables. We use ordinary standard errors. The time frame of the data is December 2007 until March 2013. Bid-ask spread has a positive relationship with CDS spreads implying that the higher the bid-ask spread (higher illiquidity), the higher the CDS spread.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities
4-) REGRESSION OF BID-ASK SPREAD STOCK LIQUIDITY WITH FIRM-LEVEL CONTROLS USING PERIOD FIXED EFFECTS
The table shows the results of the regression using period fixed effects, which includes the bid-ask spread (bas) as a proxy for stock liquidity and debt to equity ratio (levde) as a proxy for the leverage control variable. The return (r), equity value (lne), volatility (vol), operating margin (opm) and debt to equity leverage (levde) are the firm-level control variables used in this regression. We use ordinary standard errors. The time frame is from December 2007 to March 2013. The bid-ask spread has an insignificant relationship in the regression in which only period fixed effects are used.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities Intercept c 5,5414 0,0630 88,0152 0,0000 Bid Ask Spread bas -2,4678 1,3313 -1,8537 0,0638 Return r 0,0526 0,0770 0,6825 0,4949 Equity Value lne -0,1462 0,0059 -24,7765 0,0000 Volatility vol 27,0279 0,8143 33,1927 0,0000 Operating Margin opm -0,0008 0,0007 -1,1499 0,2502 Lev. Debt to Equity levde 0,0001 0,0000 4,8416 0,0000 R-squared 0,4785
F-statistic 68,8439 Prob (F-statistic) 0,0000
Period fixed effects
Periods :64 Ordinary standard errors & covariance
Cross-Sections :82
59
5-) REGRESSION OF BID-ASK SPREAD STOCK LIQUIDITY WITH FIRM-LEVEL CONTROLS USING CROSS-SECTION AND PERIOD FIXED EFFECTS
The table shows the results of the regression using both cross-section and period fixed effects, which includes the bid-ask spread (bas) as a proxy for stock liquidity and debt to equity ratio (levde) as a proxy for the leverage control variable. The return (r), equity value (lne), volatility (vol), operating margin (opm) and debt to equity leverage (levde) are the firm-level control variables used in this regression. We use ordinary standard errors. The time frame of the data used is December 2007 until March 2013. The regression using both fixed effects shows a significant positive relationship between bid-ask spread and CDS spread.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities
Periods :64 Ordinary standard errrors & covariance
Cross-Sections :82
6-) RESIDUAL GRAPH OF THE REGRESSION OF BID-ASK SPREAD STOCK LIQUIDITY USING CROSS-SECTION AND TIME FIXED EFFECTS
The graph shows the distribution of the residuals of the regression of bid-ask spread (bas) stock liquidity measure using cross-section and period fixed effects with firm-level control variables. There is a substantial improvement regarding the variance of residuals in comparison to the one in the pooled regression. However, there are still sign of heteroskedasticity.
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
500 1000 1500 2000 2500 3000 3500 4000 4500 5000
LNCDS Residuals
60
7-) WHITE TEST FOR HETEROSKEDASTICITY
In the White test for heteroskedasticity, we use square of residuals as the dependent variable and the stock liquidity measured by bid-ask spread (bas), firm-level control variables, their squares and cross-products as explanatory variables. Results of the test do not change when using the other two proxies for stock liquidity: the Amihud illiquidity measure (illiq) and turnover ratio (trnvr).
Dependent Variable: Squared Residuals Method: Panel Least Squares
The Breusch-Pagan test uses the squared residuals as dependent variable and the stock liquidity measured by bid-ask spread (bas) and firm-level controls as explanatory variables. The p-value (0,0000) of the test indicates that heteroskedasticity exists in the residuals
Dependent Variable: Squared Residuals Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities
9-) REGRESSION OF BID-ASK SPREAD STOCK LIQUIDITY USING WHITE SE
The table shows the results of the regression using both cross-section and period fixed effects, which includes the bid-ask spread (bas) as a proxy for stock liquidity and debt to equity ratio (levde) as a proxy for the leverage control variable. The return (r), equity value (lne), volatility (vol), operating margin (opm) and debt to equity leverage (levde) are the firm-level control variables used in this regression. We use White diagonal standard errors. The time frame is from December 2007 until March 2013. The bid-ask spread has a significant positive relationship with CDS spread.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities Intercept c 10,1176 0,1629 62,0971 0,0000 Bid Ask Spread bas 3,0271 1,3295 2,2769 0,0228 Return r 0,3432 0,0632 5,4303 0,0000 Equity Value lne -0,5800 0,0166 -34,8498 0,0000 Volatility vol 10,2023 0,7986 12,7751 0,0000 Operating Profit Margin opm -0,0017 0,0014 -1,1901 0,2341 Lev. Debt to Equity levde 0,0000 0,0000 4,6382 0,0000 R-squared 0,8514
F-statistic 194,7053 Prob (F-statistic) 0,0000
Cross-section and period fixed effects
Periods :64 White diagonal standard errors & covariance
Cross-Sections :82
62
10-) RESIDUAL GRAPH OF THE POOLED REGRESSION OF BID-ASK SPREAD STOCK LIQUIDITY WITH MACROECONOMIC VARIABLES
The graph shows the distribution of the residuals of the pooled regression with macroeconomic control variables. The proxy for stock liquidity is the bid-ask spread (bas). The plot indicates the presence of heteroskedasticity in the residuals of the pooled regression.
11-) RESIDUAL GRAPH OF THE REGRESSION OF BID-ASK SPREAD STOCK LIQUIDITY WITH MACRO VARIABLES USING CROSS-SECTION FIXED EFFECTS
The graph shows the distribution of the residuals of the regression of bid-ask spread (bas) stock liquidity measure using cross-section and period fixed effects with macroeconomic control variables. The variance of the residuals displays a more constant trend when using cross-section fixed effects.
-2
-1
0
1
2
3
500 1000 1500 2000 2500 3000 3500 4000 4500 5000
LNCDS Residuals
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
500 1000 1500 2000 2500 3000 3500 4000 4500 5000
LNCDS Residuals
63
12-) REGRESSION OF BID-ASK SPREAD STOCK LIQUIDITY WITH FIRM-LEVEL VARIABLES USING CROSS-SECTION AND PERIOD RANDOM EFFECTS
The table shows the results of the regression using both cross-section and period random effects, which includes the bid-ask spread (bas) as a proxy for stock liquidity and debt to equity ratio (levde) as a proxy for the leverage control variable. The return (r), equity value (lne), volatility (vol), operating margin (opm) and debt to equity leverage (levde) are the firm-level control variables used in this regression. The time frame of the data used is December 2007 until March 2013. The R2 of the regression using random effects in both dimensions is low.
Dependent Variable: CDS Spread Method: Panel EGLS (Two-way random effects)
Variables Symbol Coefficient Standard Error t-Statistic Probabilities
13-) REGRESSION OF BID-ASK SPREAD STOCK LIQUIDITY WITH THE DEBT-TO-TOTAL ASSETS LEVERAGE PROXY
The table shows the results of the regression using both cross-section and period fixed effects, which includes the bid-ask spread (bas) as a proxy for stock liquidity and debt to total assets ratio (levdta) as a proxy for the leverage control variable. The return (r), equity value (lne), volatility (vol), operating margin (opm) and debt to total assets leverage (levdta) are the firm-level control variables. We use White diagonal standard errors. The time frame of the data used is from Dec 2007 to March 2013. When using the debt to total assets ratio (levdta) proxy for leverage, the bid-ask spread still has a significant positive relationship with CDS spreads.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities
Intercept c 9,6795 0,1842 52,5410 0,0000 Bid Ask Spread bas 3,4496 1,3296 2,5946 0,0095 Return r 0,3122 0,0626 4,9860 0,0000 Equity Value lne -0,5559 0,0174 -31,9590 0,0000 Volatility vol 9,7740 0,8168 11,9664 0,0000 Operating Profit Margin opm -0,0020 0,0013 -1,5345 0,1250 Lev. Debt to Total Assets levdta 0,0046 0,0007 6,6552 0,0000 R-squared 0,8531
F-statistic 197,3801 Prob (F-statistic) 0,0000
Cross-section and period fixed effects
Periods :64
White diagonal standard errors & covariance
Cross-Sections :82
65
14-) REGRESSION OF STOCK LIQUIDITY WITH THE INTEREST COVERAGE RATIO LEVERAGE PROXY
The table shows the results of the regression using both cross-section and period fixed effects, which includes the bid-ask spread (bas) as a proxy for stock liquidity and debt to total assets ratio (levintcov) as a proxy for the leverage control variable. The return (r), equity value (lne), volatility (vol), operating margin (opm) and debt to total assets leverage (levintcov) are the firm-level control variables. We use White diagonal SEs. The time frame of data is December 2007 until March 2013. The levintcov proxy for leverage has the expected negative but insignificant relationship with CDS spread. This result does not affect the bid-ask spread’s relationship with the dependent variable.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities Intercept c 10,1225 0,1644 61,5611 0,0000 Bid Ask Spread bas 3,0521 1,3321 2,2911 0,0220 Return r 0,3396 0,0632 5,3770 0,0000 Equity Value lne -0,5796 0,0168 -34,4329 0,0000 Volatility vol 10,1977 0,8023 12,7108 0,0000 Operating Margin opm -0,0017 0,0014 -1,2005 0,2300 Lev. Interest Coverage levintcov -0,0006 0,0006 -1,0946 0,2737 R-squared 0,8511
F-statistic 194,2816 Prob (F-statistic) 0,0000
Cross-section and period fixed effects
Periods :64 White diagonal standard errors & covariance
Cross-Sections :82
66
15-) POOLED REGRESSION OF AMIHUD STOCK LIQUIDITY WITH FIRM-LEVEL CONTROL VARIABLES
The table depicts the results of the pooled regression with the Amihud illiquidity measure (illiq) as a proxy for stock liquidity and debt to equity ratio (levde) as a proxy for the leverage control variable. The equity value (lne), operating margin (opm) and debt to equity leverage (levde) are the firm-level control variables used in this regression. We use ordinary standard errors. The time frame of the data used is December 2007 until March 2013. Since there are not any fixed effects, Amihud illiquidity measure for stock liquidity shows an opposite relationship with CDS spread.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities
16-) REGRESSION OF AMIHUD STOCK LIQUIDITY WITH FIRM-LEVEL CONTROLS USING CROSS-SECTION FIXED EFFECTS
The table depicts the results of the regression using cross-section fixed effects with the Amihud illiquidity measure (illiq) as a proxy for stock liquidity and debt to equity ratio (levde) as a proxy for the leverage control variable. The equity value (lne), operating margin (opm) and debt to equity leverage (levde) are the firm-level control variables. We use ordinary standard errors. The time frame is Dec. 2007 until March 2013. When using cross-section fixed effects, Amihud illiquidity measure demonstrates the expected positive relationship between illiquidity of a stock and the CDS spread.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities Intercept c 12,6513 0,1494 84,7037 0,0000 Amihud Illiquidity illiq 126172,00 48511,25 2,6009 0,0093 Equity Value lne -0,8312 0,0156 -53,1521 0,0000 Operating Profit Margin opm 0,0080 0,0017 4,6698 0,0000 Lev. Debt to Equity levde 0,0000 0,0000 -0,4177 0,6762 R-squared 0,6985
F-statistic 140,6686 Prob (F-statistic) 0,0000
Cross-section fixed effects
Periods :64 Ordinary standard errors & covariance
Cross-Sections :82
67
17-) REGRESSION OF AMIHUD STOCK LIQUIDITY USING CROSS-SECTION FIXED EFFECTS AND WHITE CROSS-SECTION STANDARD ERRORS
The table depicts the results of the regression using cross-section fixed effects with the Amihud illiquidity measure (illiq) as a proxy for stock liquidity and debt to equity ratio (levde) as a proxy for the leverage control variable. The equity value (lne), operating margin (opm) and debt to equity leverage (levde) are the firm-level control variables used in this regression. We use White cross-section standard errors. The time frame of the data is December 2007 until March 2013. Amihud illiquidity measure for stock liquidity has a significant positive relationship with the CDS spread.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities
18-) POOLED REGRESSION OF TURNOVER RATIO STOCK LIQUIDITY WITH FIRM-LEVEL CONTROLS
The table depicts the results of the pooled regression with the turnover ratio (trnvr) as a proxy for stock liquidity and debt to equity ratio (levde) as a proxy for the leverage control variable. The return (r), equity value (lne), volatility (vol), operating margin (opm) and debt to equity leverage (levde) are the firm-level control variables. We use ordinary standard errors. The time frame of the data used is December 2007 until March 2013. A counter-intuitive positive relationship with turnover ratio measure for stock liquidity exists with CDS spread due to the pooled regression’s specifications.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities
19-) REGRESSION OF TURNOVER RATIO STOCK LIQUIDITY WITH FIRM-LEVEL CONTROLS USING CROSS-SECTION FIXED EFFECTS
The table depicts the results of the regression using cross-section fixed effects with the turnover ratio (trnvr) as a proxy for stock liquidity and debt to equity ratio (levde) as a proxy for the leverage control variable. The return (r), equity value (lne), volatility (vol), operating margin (opm) and debt to equity leverage (levde) are the firm-level control variables used in this regression. We use ordinary standard errors. The time frame of the data used is December 2007 until March 2013. The regression with cross-section fixed effects gives the expected significant relationship between turnover ratio and CDS spread. The regression implies that the higher turnover ratio (higher liquidity), the lower CDS spread.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities
20-) REGRESSION OF TURNOVER RATIO STOCK LIQUIDITY USING WHITE CROSS-SECTION STANDARD ERRORS
The table depicts the results of the regression using cross-section fixed effects with the turnover ratio (trnvr) as a proxy for stock liquidity and debt to equity ratio (levde) as a proxy for the leverage control variable. The return (r), equity value (lne), volatility (vol), operating margin (opm) and debt to equity leverage (levde) are the firm-level control variables. We use White cross-section standard errors. The time frame is Dec 2007 until March 2013. The regression using White cross-section SEs still demonstrates a significant negative relationship between turnover ratio and CDS spread as we expect.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities
21-) POOLED REGRESSION OF STOCK LIQUIDITY WITHOUT VOLATILITY AND EQUITY VALUE CONTROL VARIABLES
The table depicts the results of the pooled regression with the bid-ask spread (bas) as a proxy for stock liquidity and debt to equity ratio (levde) as a proxy for the leverage control variable. The return (r), operating margin (opm) and debt to equity leverage (levde) are the firm-level control variables. We use ordinary standard errors. The time frame of the data used is December 2007 until March 2013. The return (r) now shows a significant negative relationship with CDS spreads as the theory states while the bid-ask spread has the expected positive relationship with CDS spread.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities
Intercept c 4,7351 0,0142 332,7158 0,0000 Bid Ask Spread bas 7,3556 1,7773 4,1385 0,0000 Return r -0,6472 0,0850 -7,6126 0,0000 Operating Profit Margin opm -0,0093 0,0009 -10,7582 0,0000 Lev. Debt to Equity levde 0,0000 0,0000 2,8796 0,0004 R-squared 0,0370
F-statistic 50,2973 Prob (F-statistic) 0,0000
Pooled regression
Periods :64 Ordinary standard errors & covariance
Cross-Sections :82
22-) REGRESSION OF BID-ASK SPREAD STOCK LIQUIDITY WITHOUT VOLATILITY AND EQUITY VALUE USING CROSS-SECTION AND PERIOD FIXED EFFECTS
The table depicts the results of the regression using cross-section and period fixed effects with the bid-ask spread (bas) as a proxy for stock liquidity and debt to equity ratio (levde) as a proxy for the leverage control variable. The return (r), operating margin (opm) and debt to equity leverage (levde) are the firm-level control variables used in this regression. We use ordinary SEs. The time frame is Dec. 2007 until March 2013. The return shows a significant negative relationship with CDS spreads.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities Intercept c 4,8406 0,0207 233,5545 0,0000 Bid Ask Spread bas 8,2012 1,2062 6,7991 0,0000 Return r -0,1570 0,0522 -3,0096 0,0026 Operating Profit Margin opm -0,0172 0,0015 -11,2724 0,0000 Lev. Debt to Equity levde 0,0000 0,0000 4,3108 0,0000 R-squared 0,7645
F-statistic 111,8628 Prob (F-statistic) 0,0000
Cross-section fixed effects and period fixed effects
Periods :64 Ordinary standard errors & covariance
Cross-Sections :82
23-) REGRESSION OF BID-ASK SPREAD STOCK LIQUIDITY WITHOUT VOLATILITY AND EQUITY VALUE CONTROL VARIABLES USING WHITE STANDARD ERRORS
72
The table depicts the results of the regression using cross-section and period fixed effects with the bid-ask spread (bas) as a proxy for stock liquidity and debt to equity ratio (levde) as a proxy for the leverage control variable. The return (r), operating margin (opm) and debt to equity leverage (levde) are the firm-level control variables used in this regression. We use White diagonal standard errors. The time frame of the data used is December 2007 until March 2013. Using White diagonal SEs in the regression, the return (r) shows a significant negative relationship with CDS spreads.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities
Intercept c 4,8406 0,0297 162,8637 0,0000 Bid Ask Spread bas 8,2012 2,1359 3,8397 0,0001 Return r -0,1570 0,0710 -2,2120 0,0270 Operating Profit Margin opm -0,0172 0,0022 -7,8475 0,0000 Lev. Debt to Equity levde 0,0000 0,0000 5,8149 0,0000 R-squared 0,7645
F-statistic 111,8628 Prob (F-statistic) 0,0000
Cross-section fixed effects and period fixed effects
Periods :64 White diagonal standard errors & covariance
Cross-Sections :82
24-) REGRESSION OF TURNOVER RATIO STOCK LIQUIDITY USING THE 1-YEAR INTERBANK INT. RATE AND WHITE CROSS-SECTION STANDARD ERRORS
The table depicts the results of the regression using cross-section fixed effects with the turnover ratio (trnvr) as a proxy for stock liquidity and 1-year interbank interest rate (intbankr) as a proxy for the risk-free rate control variable. The GDP growth rate (gdp) and 1-year interbank int. rate (intbankr) are the macro control variables. We use White cross-section standard errors. The time frame used is Dec 2007 until March 2013. The turnover ratio displays a counter-intuitive result as it has a positive relationship with CDS spreads, implying that a higher stock liquidity results in higher CDS spreads.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities Intercept c 4,6189 0,0421 109,8113 0,0000 Turnover Ratio trnvr 52,8376 8,1467 6,4858 0,0000 GDP Growth Rate gdp -19,5113 2,7265 -7,1561 0,0000 Interbank Interest Rate intbankr -0,0791 0,0202 -3,9130 0,0001 R-squared 0,6384
F-statistic 108,5188 Prob (F-statistic) 0,0000
Cross-section fixed effects
Periods :64 White cross-section standard errors & covariance
Cross-Sections :82
25-) REGRESSION OF BID-ASK SPREAD STOCK LIQUIDITY USING THE 10-YEAR BOND YIELD AND WHITE CROSS-SECTION STANDARD ERRORS
73
The table depicts the results of the regression using cross-section fixed effects with the bid-ask spread (bas) as a proxy for stock liquidity and 10-year bond yield (fiveyld) as a proxy for the risk-free rate control variable. The GDP growth rate (gdp) and 10-yr bond yield (tenyld) are the macro-level control variables used in this regression. We use White cross-section standard errors. The time frame of the data used is Dec2007 until March 2013. When using the ten-year bond yield, the bid-ask spread still displays a positive relationship with the CDS spreads.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard
Error t-Statistic Probabilities Intercept c 4,9871 0,0834 59,7847 0,0000 Bid-Ask Spread bas 12,3237 3,1372 3,9282 0,0001 GDP Growth Rate gdp -18,6407 2,3652 -7,8813 0,0000 10-year Government bond yield tenyld -0,1189 0,0270 -4,4065 0,0000 R-squared 0,6409
F-statistic 109,7414 Prob (F-statistic) 0,0000
Cross-Section Fixed Effects
Periods :64 White cross-section standard errors & covariance
Cross-Sections :82
26-) POOLED REGRESSION OF BID-ASK SPREAD STOCK LIQUIDITY WITH MACROECONOMIC CONTROL VARIABLES
The table depicts the results of the pooled regression with the bid-ask spread (bas) as a proxy for stock liquidity and 1-year interbank rate (intbankr) as a proxy for the risk-free rate control variable. The GDP growth rate (gdp) and 1-year interbank rate (intbankr) are the macro-level control variables used in this regression. We use ordinary standard errors. The time frame of the data used is Dec 2007 until March 2013. The bid-ask spread has an expected positive relationship with the dependent variable.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities
Intercept c 4,7341 0,0140 338,3695 0,0000 Bid Ask Spread bas 6,7159 1,7129 3,9209 0,0001 GDP Growth Rate gdp -21,0662 0,8684 -24,2582 0,0000 1-year Interbank rate intbankr -0,0524 0,0051 -10,3354 0,0000 R-squared 0,1047
F-statistic 204,4443 Prob (F-statistic) 0,0000
Pooled Regression
Periods :64 Ordinary standard errors & covariance
Cross-Sections :82
27-) REGRESSION OF STOCK LIQUIDITY WITH MACROECONOMIC CONTROL VARIABLES USING CROSS-SECTION FIXED EFFECTS
74
The table depicts the results of the regression using cross-section fixed effects with the bid-ask spread (bas) as a proxy for stock liquidity and 1-year interbank rate (intbankr) as a proxy for the risk-free rate control variable. The GDP growth rate (gdp) and 1-year interbank rate (intbankr) are the macro-level control variables used in this regression. We use ordinary standard errors. The time frame of the data used is December 2007 until March 2013. When using cross-section fixed effects, the explanatory power of bid-ask spread on CDS spread increases in comparison to the pooled regression.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities
Intercept c 4,7125 0,0095 495,9753 0,0000 Bid Ask Spread bas 13,0017 1,5083 8,6202 0,0000 GDP Growth Rate gdp -19,8148 0,5779 -34,2876 0,0000 1-year Interbank rate intbankr -0,0480 0,0034 -14,1356 0,0000 R-squared 0,6217
F-statistic 100,9973 Prob (F-statistic) 0,0000
Cross-Section Fixed Effects
Periods :64 Ordinary standard errors & covariance
Cross-Sections :82
28-) REGRESSION OF BID-ASK SPREAD STOCK LIQUITIY WITH MACROECONOMIC CONTROL VARIABLES USING WHITE CROSS-SECTION STANDARD ERRORS
The table depicts the results of the regression using cross-section fixed effects with the bid-ask spread (bas) as a proxy for stock liquidity and 1-year interbank rate (intbankr) as a proxy for the risk-free rate control variable. The GDP growth rate (gdp) and 1-year interbank rate (intbankr) are the macro-level control variables. We use White cross-section standard errors. The time frame of the data used is December 2007 until March 2013. Bid-ask spread has a significant positive relationship with CDS spread when using White cross-section standard errors.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard Error t-Statistic Probabilities
Intercept c 4,7125 0,0390 120,6527 0,0000 Bid Ask Spread bas 13,0017 3,3867 3,8391 0,0001 GDP Growth Rate gdp -19,8148 2,6648 -7,4359 0,0000 1-year Interbank rate intbankr -0,0480 0,0206 -2,3308 0,0198 R-squared 0,6217
F-statistic 100,9973 Prob (F-statistic) 0,0000
Cross-Section Fixed Effects
Periods :64 White cross-section standard errors & covariance
Cross-Sections :82
29-) POOLED REGRESSION OF BID-ASK SPREAD STOCK LIQUIDITY USING THE 5-YEAR BOND YIELD PROXY FOR RISK-FREE RATE
75
The table depicts the results of the pooled regression with the bid-ask spread (bas) as a proxy for stock liquidity and 5-year bond yield (fiveyld) as a proxy for the risk-free rate control variable. The GDP growth rate (gdp) and 5-year bond yield (fiveyld) are the macro-level control variables used in this regression. We use ordinary standard errors. The time frame of the data used is December 2007 until March 2013. When using 5-yr bond yield proxy for risk-free rate, the bid ask spread continues to have a significant positive relationship with CDS spread.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard
Error t-Statistic Probabilities
Intercept c 4,8582 0,0172 283,1067 0,0000 Bid-Ask Spread bas 6,4967 1,6898 3,8448 0,0001 GDP Growth Rate gdp -20,3585 0,8053 -25,2798 0,0000 5-year Government bond yield fiveyld -0,1055 0,0066 -15,9327 0,0000 R-squared 0,1287
F-statistic 258,0934 Prob (F-statistic) 0,0000
Pooled Regression
Periods :64 Ordinary standard errors & covariance
Cross-Sections :82
30-) REGRESSION OF BID-ASK SPREAD STOCK LIQUIDITY USING THE 5-YEAR BOND YIELD AND CROSS-SECTION FIXED EFFECTS
The table depicts the results of the regression using cross-section fixed effects with the bid-ask spread (bas) as a proxy for stock liquidity and 5-year bond yield (fiveyld) as a proxy for the risk-free rate control variable. The GDP growth rate (gdp) and 5-yr bond yield (fiveyld) are the macro-level control variables. We use ordinary standard errors. The time frame of the data used is December 2007 until March 2013.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard
Error t-Statistic Probabilities Intercept c 4,8469 0,0113 427,3549 0,0000 Bid-Ask Spread bas 12,1310 1,4565 8,3288 0,0000 GDP Growth Rate gdp -19,3418 0,5227 -37,0032 0,0000 5-year Government bond yield fiveyld -0,1051 0,0043 -24,3549 0,0000 R-squared 0,6475
F-statistic 112,9133 Prob (F-statistic) 0,0000
Cross-Section Fixed Effects
Periods :64 Ordinary standard errors & covariance
Cross-Sections :82
31-) REGRESSION OF BID-ASK SPREAD STOCK LIQUIDITY USING THE 5-YEAR BOND YIELD AND WHITE CROSS-SECTION STANDARD ERRORS
76
The table depicts the results of the regression using cross-section fixed effects with the bid-ask spread (bas) as a proxy for stock liquidity and 5-year bond yield (fiveyld) as a proxy for the risk-free rate control variable. The GDP growth rate (gdp) and 5-year bond yield (fiveyld) are the macro-level control variables. We use White cross-section SEs. The time frame is Dec. 2007 until March 2013. By controlling the risk-free rate measured by the 5-year bond yield in the regression using White cross-section SEs, we obtain a significant positive relationship between bid-ask spread and CDS spread.
Dependent Variable: CDS Spread Method: Panel Least Squares
Variables Symbol Coefficient Standard
Error t-Statistic Probabilities Intercept c 4,8469 0,0490 98,9203 0,0000 Bid-Ask Spread bas 12,1310 3,0844 3,9331 0,0001 GDP Growth Rate gdp -19,3418 2,3830 -8,1165 0,0000 5-year Government bond yield fiveyld -0,1051 0,0220 -4,7783 0,0000 R-squared 0,6475
F-statistic 112,9133 Prob (F-statistic) 0,0000
Cross-Section Fixed Effects
Periods :64 White cross-section standard errors & covariance
Cross-Sections :82
32-) POOLED REGRESSION OF AMIHUD STOCK LIQUIDITY WITH MACROECONOMIC CONTROL VARIABLES
The table depicts the results of the pooled regression with the Amihud illiquidity measure (illiq) as a proxy for stock liquidity and 1-year interbank rate (intbankr) as a proxy for the risk-free rate control variable. The GDP growth rate (gdp) and 1-year interbank rate (intbankr) are the macro-level control variables used in this regression. We use ordinary standard errors. The time frame of the data used is December 2007 until March 2013. The pooled regression with macro control variables gives a counter-intuitive negative relationship between Amihud illiquidity measure and CDS spread.
Dependent Variable: CDS Spread Method: Panel Least Squares
33-) REGRESSION OF AMIHUD STOCK LIQUIDITY WITH MACROECONOMIC CONTROL VARIABLES USING CROSS-SECTION FIXED EFFECTS
77
The table depicts the results of the regression using cross-section fixed effects with the Amihud illiquidity measure (illiq) as a proxy for stock liquidity and 1-year interbank rate (intbankr) as a proxy for the risk-free rate control variable. The GDP growth rate (gdp) and 1-year interbank rate (intbankr) are the macro-level control variables used. We use ordinary SEs. The time frame is Dec. 2007 until March 2013. When cross-section fixed effects are used in the regression with macro control variables, Amihud illiquidity measure shows the expected positive relationship with a significance of 3.18%.
Dependent Variable: CDS Spread Method: Panel Least Squares
34-) REGRESSION OF AMIHUD STOCK LIQUIDITY WITH MACROECONOMIC CONTROL VARIABLES USING WHITE CROSS-SECTION STANDARD ERRORS
The table depicts the results of the regression using cross-section fixed effects with the Amihud illiquidity measure (illiq) as a proxy for stock liquidity and 1-year interbank rate (intbankr) as a proxy for the risk-free rate control variable. The GDP growth rate (gdp) and 1-year interbank rate (intbankr) are the macro-level control variables. We use White cross-section ordinary SEs. The time frame of the data is Dec. 2007 until March 2013. When using White cross-section SEs, Amihud illiquidity measure has the expected positive relationship with CDS spread but a significance of 5.51%, slightly higher than 5% significance level.
Dependent Variable: CDS Spread Method: Panel Least Squares
Periods :64 White cross-section standard errors & covariance
Cross-Sections :82
78
35-) THE LIST OF 82 FIRMS USED IN THE SAMPLE:
* Company Name Country Company Name Country 1 Aktiebolaget Volvo Sweden 42 Next PLC UK 2 Akzo Nobel N.V. Netherlands 43 PPR France 3 Alstom France 44 Sabmiller PLC UK 4 Anglo American PLC UK 45 Suedzucker AG Mannheim Germany 5 Astrazeneca PLC UK 46 Tate & Lyle PLC UK 6 Atlantia S.P.A. Italy 47 Tesco PLC UK 7 BAE Systems PLC UK 48 Unilever UK 8 BASF SE UK 49 BP PLC UK 9 Bayer Aktiengesellschaft Germany 50 Centrica PLC UK
10 Bayerische Motoren Werke AG Germany 51 E.ON AG Germany 11 Bouygues France 52 Electricite de France (EDF) France 12 Compagnie de Saint-Gobain France 53 EnBW Energie Baden-Wuerttemberg AG Germany 13 Compagnie Financiere Michelin France 54 ENEL S.P.A. Italy 14 Daimler AG Germany 55 ENI S.P.A. Italy 15 Holcim Ltd Switzerland 56 Fortum Oyj Finland 16 Lanxess AG Germany 57 Gas Natural SDG S.A. Spain 17 Linde AG Germany 58 GDF Suez France 18 Rentokil Initial PLC UK 59 Iberdrola S.A. Spain 19 Rolls-Royce PLC UK 60 National Grid PLC UK 20 Sanofi France 61 Royal Dutch Shell PLC Netherlands 21 Siemens AG Germany 62 RWE AG Germany 22 Solvay Belgium 63 Total SA France 23 Valeo France 64 United Utilities PLC UK 24 Vinci France 65 Veolia Environment France 25 Volkswagen AG Germany 66 Aegon N.V. Netherlands 26 Xstrata PLC UK 67 British Telecommunications PLC UK 27 Accor France 68 Deutsche Telekom AG Germany 28 Aktiebolaget Electrolux Sweden 69 France Telecom France 29 Anheuser-Busch InBev Belgium 70 Koninklijke KPN N.V. Netherlands 30 British American Tobacco PLC UK 71 Pearson PLC UK 31 Carrefour France 72 Publicis Groupe SA France 32 Compass Group UK 73 STMicroelectronics N.V. France 33 Diageo PLC UK 74 Telecom Italia S.P.A. Italy 34 Experian Finance PLC UK 75 Telefonaktiebolaget LM Ericsson Sweden 35 Imperial Tobacco Group PLC UK 76 Telekom Austria AG Austria 36 Kingfisher PLC UK 77 Telenor ASA Norway 37 Koninklijke Ahold N.V. Netherlands 78 TeliaSonera Aktiebolaget Sweden 38 LVMH France 79 Vivendi France 39 Marks and Spencer PLC UK 80 Vodafone Group PLC UK 40 Metro AG Germany 81 Wolters Kluwer N.V. Netherlands 41 Nestle S.A. Switzerland 82 WPP 2005 Limited UK
Number of Companies per Country
United Kingdom 27 Sweden 3 France 19 Belgium 2 Germany 13 Switzerland 2 Netherlands 6 Norway 1 Italy 4 Finland 1 Spain 3 Austria 1 *The companies are selected from Markit iTraxx Europe Reference Portfolio